def test_qc_maf():
random = RandomState(0)
X = random.randint(0, 3, size=(100, 10))
assert_allclose(
compute_maf(X), [0.49, 0.49, 0.445, 0.495, 0.5, 0.45, 0.48, 0.48, 0.47, 0.435]
)
def test_compute_maf_numpy():
random = RandomState(0)
X = random.randint(0, 3, size=(100, 10))
maf = compute_maf(X)
assert_allclose(
maf, [0.49, 0.49, 0.445, 0.495, 0.5, 0.45, 0.48, 0.48, 0.47, 0.435])
def maf_filter(self, thresh =0.05):
from limix.qc import compute_maf
print("Start SNP filtering....(MAF > %s)" %thresh)
maf_list = compute_maf(self.geno_matrix.T.values)
self.geno_matrix = self.geno_matrix[np.array(maf_list) > thresh]
self.SNPinfo = self.SNPinfo[np.array(maf_list) > thresh]
print('%s variations used to perform GWAS.' %self.SNPinfo.shape[0])
def test_compute_maf_dask_array():
random = RandomState(0)
X = da.from_array(random.randint(0, 3, size=(100, 10)), chunks=2)
maf = compute_maf(X)
assert_(isinstance(maf, ndarray))
assert_allclose(
maf, [0.49, 0.49, 0.445, 0.495, 0.5, 0.45, 0.48, 0.48, 0.47, 0.435])
def test_compute_maf_dataframe():
random = RandomState(0)
X = random.randint(0, 3, size=(100, 10))
columns = [f"snp{i}" for i in range(X.shape[1])]
maf = compute_maf(DataFrame(X, columns=columns))
assert_(isinstance(maf, Series))
assert_(maf.name == "maf")
assert_(array_equal(maf.index, columns))
assert_allclose(
maf, [0.49, 0.49, 0.445, 0.495, 0.5, 0.45, 0.48, 0.48, 0.47, 0.435])
def test_compute_maf_dataarray():
random = RandomState(0)
X = random.randint(0, 3, size=(100, 10))
samples = [f"snp{i}" for i in range(X.shape[0])]
candidates = [f"snp{i}" for i in range(X.shape[1])]
X = xr.DataArray(
X,
dims=["sample", "candidate"],
coords={
"sample": samples,
"candidate": candidates
},
)
maf = compute_maf(X)
assert_(isinstance(maf, xr.DataArray))
assert_(maf.name == "maf")
assert_(array_equal(maf.candidate, candidates))
assert_allclose(
maf, [0.49, 0.49, 0.445, 0.495, 0.5, 0.45, 0.48, 0.48, 0.47, 0.435])
